Title: “Statistics Project phase 1” Group: 11 Section: 001 Team members:Sujata,Vidya output: html_notebook
This project represents our (Vidya and Sujata) own work and we have adhered to St. Clair College’s Academic Integrity policies in completing this project.
R version 4.2.2
RStudio 2022.12.0+353 “Elsbeth Geranium” Release (7d165dcfc1b6d300eb247738db2c7076234f6ef0, 2022-12-03) for Windows Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) RStudio/2022.12.0+353 Chrome/102.0.5005.167 Electron/19.1.3 Safari/537.36
Packages: ggplot2,dplyr,plotly Hence, before using these library we need to install these packages install.packages(ggplot2) install.packages(dplyr) install.packages(plotly) install.packages(patchwork)
This analysis will be based on the dataset that contains information of 1465 Amazon products, ratings and reviews, as reported on the company’s official website. https://drive.google.com/drive/folders/1O5ybNrX_kAPskp_gVVlxbvNIE6GtPbxV
Data preprocessing:- As there are lot of categories we need to reduce them into main categories. To acheive this we have used =LEFT(C2, FIND(“|”, C2, FIND(“|”, C2)+1)-1) formula to create a new column called main_category which is used to vizualize the sales data in each category.
#importing a ggplot2 library
library(ggplot2)
library(plotly)##
## Attaching package: 'plotly'## The following object is masked from 'package:ggplot2':
##
## last_plot## The following object is masked from 'package:stats':
##
## filter## The following object is masked from 'package:graphics':
##
## layoutlibrary(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union#reading a csv file from excel
amazon<-read.csv("C:/Users/vidya/OneDrive/Desktop/Stats/projectphase1/a2.csv")
#Data preprocesing
amazon$discounted_price <- gsub("[^0-9]", "", amazon$discounted_price)
amazon$actual_price <- gsub("[^0-9]", "", amazon$actual_price)
amazon$rating_count <- gsub("[^0-9]", "", amazon$rating_count)
amazon$discount_percentage <- gsub("[^0-9]", "", amazon$discount_percentage)#1
#@author Vidya
#@Editor Sujata
#@returns A histogram which displays the total number of ratings
#@sample run load variable and execute ggplot snippet
variable <- amazon$rating
ggplot(data.frame(variable), aes(variable)) +
geom_bar(stat = "count",fill = "cyan4",color = "white") +
labs(x = "RATING",y = "FREQUENCY",
title = "RANGE OF RATING FROM 1 TO 5",
caption = "Bar graph showing the distribution of 'Rating' over the range 1 to 5. People have given 4.1 rating 250 times, followed by 4.2 & 4.3 around 225 times ")+
theme(plot.title = element_text(face = "bold", hjust = 0.5),
plot.caption = element_text(hjust = 0.5, color = "gray27"),
panel.grid = element_blank(),
panel.background = element_rect(fill = "white"),
axis.text.x = element_text(color = "darkslategray4"),
axis.text.y = element_text(color = "darkslategray4"))
#1
#@author Vidya
#@Editor Sujata
#@returns A density plot which displays the density of ratings
#@sample run load variable and execute ggplot snippet
#there is NA value in the rating(to avoid repetition used 'rate' instead of 'rating' )
rate<-as.numeric(amazon$rating)## Warning: NAs introduced by coercion# checking the datatype of rate
typeof(rate)## [1] "double"#replacing the NA value with 0
rate<-replace(rate,is.na(rate),0)
table(is.na(rate))##
## FALSE
## 1465#Density displaying rating
ggplot(amazon, aes(x=rate)) +
geom_density()+
labs(y= "Density",x="Rating",
title = "HIGHEST DENSITY OF RATING",
caption = "The Graph Shows Rating as a Density Plot") +
theme(plot.title = element_text(face = "bold",
hjust = 0.5,
color = "pink3"),
panel.background = element_rect(fill = "white"),
panel.grid.minor = element_line(color = "pink3"),
plot.caption = element_text(color = "pink3",
hjust = 0.5))
#2
#@author Vidya
#@Editor Sujata
#@returns A pie chart which displays the highest selling product and its share
#@sample run load variable and execute ggplot snippet
# calculate percentages for each main_category
prop_table <-amazon %>% count(main_category) %>% mutate(props = n / sum(n))
ggplot(prop_table) +
geom_bar(aes(x = "", y = props, fill = main_category),width = 1,stat = "identity") +
coord_polar("y", start = 0) +
labs(x = "",y = "",
title = "Top selling product categories and their percentage share",
caption = "Different Product Categories displayed as a Pie Diagram") +
theme(plot.title = element_text(face = "bold",hjust = -0.1,color = "dodgerblue4"),
plot.caption = element_text(hjust = 0.5, color = "dodgerblue4"),
panel.grid = element_blank(),
panel.background = element_rect(fill = "white"),
legend.text = element_text(color = "dodgerblue4"),
legend.position = "right",
legend.title = element_blank()) +
guides(fill = guide_legend(title = "main category")) +
scale_fill_discrete(labels = paste(prop_table$main_category,
scales::percent(prop_table$props),sep = " - "))
#2
#@author Vidya
#@Editor Sujata
#@returns A bar plot which displays the total product sold in each category
#@sample run load variable and execute ggplot snippet
#bar chart of category and sales in each category
ggplot(data = amazon) +
geom_bar(mapping = aes(x = main_category),
fill = 'steelblue') +
labs(x = "",y = "",title = "NUMBER OF RATINGS IN EACH CATEGORY",
caption = "Bar Graph Depicting Total Count of Rating in Each Category") +
theme(plot.title = element_text(face = "bold",
hjust = 0.5,
color = "steelblue4"),
plot.caption = element_text(hjust = 0.5,
color = "steelblue4"),
axis.text.x = element_text(angle = 45,
hjust = 1,
color = "steelblue2"),
axis.text.y = element_text(color = "steelblue2"),
panel.grid.minor = element_line(color = "pink3"),
panel.background = element_rect(fill = "white"))
#3
#@author Vidya
#@Editor Sujata
#@returns A point plot which displays discounted percentage and rating for each category
#@sample run load variable and execute ggplot snippet
#enhance the grapgh to full screen to view all percentage
ggplot(data = amazon,mapping = aes(x = discount_percentage ,
y = rating,
color = main_category)) +
geom_point() +
theme(axis.text.x = element_text(angle = 90,
vjust = 0.5,
hjust = 1)) +
scale_y_discrete(breaks = seq(0, 5, 0.5)) +
labs(x = "DISCOUNT PERCENTAGE",y = "RATING",
title = "DISCOUNT PERCENTAGE INFLUENCING RATING OF EACH PRODUCT CATEGORY",
caption = "Point Graph Showing the Relationship between Discount Percentage & Each Product category") +
theme(plot.title = element_text(face = "bold",
hjust = 0.5,
color = "hotpink4"),
plot.caption = element_text(hjust = 0.5 ,
color = "hotpink4"),
panel.grid = element_blank(),
panel.background = element_rect(fill = "white"),
axis.text.x = element_text(size=5,
angle = 90,
color = "steelblue"),
axis.text.y = element_text(color = "hotpink4"),
legend.text = element_text(color = "hotpink4"))
#3
#@author Sujata
#@Editor Vidya
#@returns A scatter plot which displays category and its rating
#@sample run load variable and execute ggplot snippet
#enhance the grapgh to full screen to view all percentage
ggplot(data = amazon) +
geom_bin2d(mapping = aes(x = main_category,
y = rating))+
labs(x = "", y = "",
title = "TOTAL COUNT IN PRODUCT CATEGORY",
caption = "Scatter Plot made with Product Category Against the Rating Count") +
theme(plot.title = element_text(face = "bold",
hjust = 0.5,
color = "steelblue4"),
plot.caption = element_text(hjust = 0.5,
color = "steelblue4"),
axis.text.x = element_text(angle = 90,
vjust = 0.5,
hjust=1,
color = "steelblue2"),
axis.text.y = element_text(color = "steelblue2"),
panel.background = element_rect(fill = "white"),
legend.position = "top")
#4
#@author Vidya
#@Editor Sujata
#@returns A bar plot which displays the product category filled with rating
#@sample run load variable and execute ggplot snippet
ggplot(data = amazon) +
geom_bar(mapping = aes(x = main_category,
fill = rating)) +
labs(x = "",y = "",
title = "PRODUCT CATEGORY ACCORDING THEIR RATING COUNT",
caption = "Bar Chart Displaying Relationship between Product Category & Rating Coungt Filled with Rating") +
theme(axis.text.x = element_text(angle = 45,
hjust = 1,
color = "slateblue1"),
axis.text.y = element_text(color = "slateblue1"),
plot.title = element_text(face = "bold",
hjust = 0.5,
color = "slateblue4"),
plot.caption = element_text(hjust = 0.5,
color = "slateblue4"),
legend.text = element_text(color = "slateblue4"),
panel.grid.major = element_line(color = "seashell2"),
panel.background = element_rect(fill = "white"))
#4
#@author Vidya
#@Editor Sujata
#@returns A bar plot which displays the max discount percentage in each category
#@sample run load variable and execute ggplot snippet
max_discount <-amazon %>% group_by(main_category) %>% summarise(max_discount = max(discount_percentage, na.rm = TRUE))
ggplot(max_discount, mapping = aes(x = main_category, y = max_discount)) +
geom_bar(stat = "identity",fill = "tan") +
labs(x = " ",title = "Maximum discount percentage in each product category",
caption = "Bar Graph of Product Category with Maximum Discount displaying relation between two variables") +
theme(axis.text.x = element_text(angle = 45,vjust = 0.8,hjust = 1),
plot.title = element_text(face = "bold",hjust = 0.5,color = "tan4"),
plot.caption = element_text(hjust = 0.5, color = "tan4"),
axis.text.y = element_text(color = "tan4"),
panel.grid = element_blank(),
panel.background = element_rect(fill = "white"))
#5
#@author Sujata
#@Editor Vidya
#@returns A point plot which displays the price between product category on rating
#@sample run load variable and execute ggplot snippet
t <- ggplot(amazon,
mapping = aes(discounted_price, actual_price,
color = main_category)) +
geom_point()
t + facet_grid(cols = vars(rating),
rows = vars(main_category)) +
labs(title = "PRICE AMONGST PRODUCT CATEGORIES REFLECTING ON THE RATING",
caption = "Difference between Discounted Price & Actual Price with respect to the Product Category and it's Rating") +
theme(
plot.title = element_text(
face = "bold",
hjust = 0.5,
color = "thistle4"
),
plot.caption = element_text(
hjust = 0.5,
color = "thistle4",
face = "italic"
),
axis.text.x = element_blank(),
axis.text.y = element_blank(),
strip.text = element_text(size = 6),
legend.position = "bottom"
)
#6
#@author Vidya
#@Editor Sujata
#@returns A bar interactive plot which displays rating and sales per each category
#@sample run load variable and execute ggplot snippet
#Hover over each bars to display sales counts and rating for each category
vi <- ggplot(amazon, aes(main_category)) +
geom_bar(aes(fill = rating)) +
labs(
x = " ",
title = "Interactive plot
displaying the sales figures along with rating for each category",
caption = " bar graph interactive"
) +
theme(
axis.text.x = element_text(angle = 45, hjust = 1),
plot.title = element_text(
face = "bold",
hjust = 0.5,
color = "tan4"
),
plot.caption = element_text(hjust = 0.5, color = "tan4"),
axis.text.y = element_text(color = "tan4"),
panel.grid = element_blank(),
panel.background = element_rect(fill = "white")
)
ggplotly(vi)REFERENCES:- [1]“Data Mentor,” DataMentor, Nov. 23, 2017. https://www.datamentor.io/r-programming/box-plot/ [2]“Colors in R.” Available: http://www.stat.columbia.edu/~tzheng/files/Rcolor.pdf [3]“ggplot2 package - RDocumentation,” www.rdocumentation.org. https://www.rdocumentation.org/packages/ggplot2/versions/3.4.1 [4]“Package ‘ggplot2’ Title Create Elegant Data Visualisations Using the Grammar of Graphics,” 2020. Available: https://cran.r-project.org/web/packages/ggplot2/ggplot2.pdf [5] A. Misiuro, “What is Data Visualization | Top Data Visualization Tools,” Synder blog, Jun. 17, 2022. https://synder.com/blog/what-is-data-visualization/
Answers to the questions given in project#1 document:
In what ways do you think data visualization is important to understanding a data set? Ans: It allows us to examine and find patterns, trends, and connections within the data that might not be immediately apparent by looking at absolute numbers or statistics. Without visualisations, stories cannot be told in the age of ever-growing data volumes. No matter how complicated the case is, the right representation could be able to bring everyone together. Visualization also helps use to summarize large amounts of data and compare it with other datasets as and when needed. This can be done to compare trends, patterns and thus get better insights. Data visualization also helps with identifying the outliers and anomalies which you might further need to understand and investigate or ignore in some situations. Visualization helps us turn numbers into useful knowledge. It makes data easier to perceive, interact with, and comprehend for users.
In what ways do you think data visualization is important to communicating important aspects of a data set? Ans: Data visualisation is a vital method for highlighting important aspects of a data set. Visualization can be used to communicate ideas and conclusions to a broad audience in a clear and concise manner. Data visualization helps one understand the data without having to look at all the numbers and other unnecessary information present in the dataset. If done properly, it also helps the viewer get insights on the data which can be used to make decisions by predicting what trend a particular variable is following. For the ones who don’t understand the minute details of the data or is not interested in doing so, visualization is the perfect way to avoid this and still be able to interact with the data. Datasets can be overly complicated and difficult to understand at times. Visualizations can help simplify this and present the data in a much more simple and understandable format. All in all, data visualization is a great tool to interact with the data while understanding it in the simplest manner.
What role does your integrity as an analyst play when creating a data visualization for communicating results to others? Ans: All data analysis, includes data visualisation, must prioritise integrity. As an analyst the analysis that one makes should be accurate and must display only facts. The truth of the data should not be manipulated or played with. The visualization must only convey genuine information about the data. The analyst’s integrity plays a vital role in doing so. A good analyst should also be transparent with how he/she got to the analysis that they did thus addressing the source and methods used for the analysis. The objective of the analyst also plays an important role in visualization. For example, the data analyst can show a sales record as a company facing loss because of a certain product or show it as how the company can maximize it’s profits but excluding a certain product and stocking up a particular one more than what the company is doing presently, or maybe both. The analyst should be responsive about the visualization made, meaning that if any concerns or queries are raised about the visualization, the analyst must be able to answer and resolve them with ease. To summarise, integrity of an analyst plays a very important role in creating data visualization to help communicate the results to others.
How many variables do you think you can successfully represent in a visualization? What happens when you exceed this number? Ans: How many variables may be accurately depicted in a visualization depends on the type of plot, the complexity of the dataset, and the information processing ability of the viewer. A recommended guideline is to keep a visualization’s variable count to no more than 5 or 6. In general, it becomes harder to fully understand the narrative of a visualization the more elements there are. This enables the viewer to focus on the main aspects of the content and helps to minimize information overload. As you go beyond this limit, visual clutter could appear, making it difficult to see connections and patterns in the data. This makes it difficult to interpret the data making it very confusing. This can also lead to a loss of context because it’s possible for important information to get buried in a sea of data. As you go beyond the limit, the variables may not be distinguishable anymore. Due of this, the viewer may not receive the intended insights and discoveries from the visualisation, thus becoming misleading and losing the purpose of visualization in the process.
Contribution:- As it is a team work we both have done equal amount of work in every part of the project. The dataset selection, the formation of the problem statement, the creation of graphs, answers to the 4 questions and the entire summary was done by the both of us equally and there has been equal contribution to the project at every step from both the members.